The firms behind natural disaster models are now predicting armed conflict, and one called the Iran war six weeks before it started.
The same analytical tools that help insurers price hurricanes and wildfires are being retooled for something far less predictable: war. As military conflicts multiply across the globe, Wall Street’s old risk playbook is breaking down, and a new one is being written in real time.
Since 2008, the number of countries in active external conflict has nearly doubled to over 100. The cost of global violence now runs close to $22 trillion a year, more than 10% of world GDP, according to the Institute for Economics and Peace.
What happened
Risk consultancy Verisk Maplecroft, better known for catastrophe modeling for insurers and cat-bond investors, has released two new tools aimed squarely at geopolitical instability.
The first, its Predictive War Index, uses machine learning to estimate the probability of war breaking out in a given country over the next 12 months. Trained on political, economic, and social data from 1995 to 2022, the model doesn’t include the current Iran conflict in its dataset, yet back-testing found it would have flagged a 66% probability of war in Iran as early as January, roughly six weeks before fighting began on February 28.
A separate Verisk model, launched in late 2023, has correctly called six out of seven government collapses since then, including Bashar al-Assad’s fall in Syria and the sudden removal of Venezuela’s Nicolás Maduro in January.
Meanwhile, the RAND Corporation is running an AI model that puts concrete probability figures on scenarios like regime change. As of mid-May, that model placed a 20% chance on Iran’s government not surviving past 2027.
On the shipping front, marine war risk premiums in the Strait of Hormuz jumped to as high as 1% of a vessel’s value per voyage after the Iran conflict began, up from a fraction of a percent previously, according to Moody’s. The US and Iran announced an interim deal to reopen the strait over the weekend, with formal talks scheduled in Switzerland on June 19, though a final text has not been released.
Market reaction
The Strait of Hormuz disruption sent marine insurance costs surging and flagged the fragility of global shipping chokepoints. Oil pricing models, mortgage rate assumptions, and supply chain risk calculations are all being stress-tested. War, as Citi’s Krishan Sharma put it, doesn’t behave like a standard deviation move in a normal distribution – “it changes the distribution entirely.”
Why it matters for traders
Both Citigroup and Morgan Stanley have sounded the alarm on legacy risk frameworks. Citi warns against “rear-view mirror” models built on historical data. Morgan Stanley says the entire approach to geopolitical risk needs a rethink. Allianz’s latest risk survey found war has now overtaken civil unrest as the political threat companies fear most when shopping for insurance.
The new models aren’t just descriptive; they’re designed to show how specific actions (sanctions, diplomacy, military posture) would shift the probability of various outcomes. For portfolio managers and underwriters, that’s a fundamentally different kind of input than anything previously available.
What to watch
The June 19 Switzerland talks will be a key signal for Hormuz premium direction. Broader adoption of predictive war models by major insurers and banks could reshape how geopolitical risk is priced across asset classes, from sovereign debt to commodity futures. With Morgan Stanley describing the current moment as the twilight of “globalization-driven efficiency,” this shift in risk modeling may be as structural as it gets.
Stay ahead of every market-moving headline with QuoMarkets.
